1. About JuiceFS Caching On a host, warmed cache is placed directly on the host. In a cluster, there are two levels of cache: Worker, which provides cache shared at the cluster level Fuse, which provides cache at the current node level only 2. Warming Data with the JuiceFS Client Note that warming at the Fuse layer is only effective for the current node; if you need to warm the whole cluster, you must warm at the Worker layer.

1. Data Delivery Based on Object Storage As shown above, model R&D mainly involves three sub-platforms: Data platform The data platform is mainly responsible for data-related management, such as data ingestion and data processing, ultimately producing the data needed for training. The data platform stores raw data in object storage. During processing, it fetches data from object storage, processes it, and stores the result back in object storage.

1. Making Ops Copilot an Ops Coilot In September 2023, I wrote a version of Ops Copilot, and published an article about it, I’m Writing a Copilot for My Ops Tool. The effect it achieved was like this: 1 2 3 4 5 6 7 8 9 10 Opscli> 打

1. What Is MLOps MLOps is short for Machine Learning Operations, and it describes the standardization and engineering of the entire lifecycle around model development. MLOps includes the following key steps: Data management: storing, accessing, cleaning, and transforming data Model development: algorithm development and model construction Model training and tuning: training models on data, adjusting hyperparameters to optimize the model, and fine-tuning models Model evaluation: testing a model’s accuracy, generalization ability, and performance metrics Model deployment: deploying the model to the target environment and converting the model Model monitoring: monitoring the model’s behavior, performance degradation, data drift, and robustness Another similar concept is ModelOps.

1. Data Parallelism Training steps: The master device loads the model and copies the model parameters to every worker device The master device splits the training data along the batch dimension and passes each batch to every worker device Each worker device performs training The master device aggregates the gradients from each worker device and updates the model parameters The master device broadcasts the model parameters to every worker device, preparing for the next batch of training Core idea: